A buyer does not pay for last year's revenue. He pays for the portion of revenue that will return next year with a reasonable degree of certainty, without having to be sold for again each time. Contract value, renewal rhythm, notice periods and the question of whether a customer can leave without switching costs: that determines whether revenue counts as recurring or is treated as one-off in the valuation.
The shift that AI is setting in motion touches this directly. Part of the work that used to cost hours to serve a customer is now done by software, partly with human oversight that approves or rejects, partly fully automatically. That does not change the revenue itself, but it does change the cost of retaining that revenue. Recurring revenue running on a process with little manual work weighs more heavily than recurring revenue that only persists thanks to the attention of one account manager.
A buyer does not look at the label "subscription" or "contract"; he looks at what happens when no one is watching. Questions that recur in due diligence:
That last point is where the measurement changes. A business in which invoicing, renewal follow-up and basic service largely run through systems shows a buyer that recurring revenue does not depend on the continuity of individuals. A business in which that work still lies entirely with people must compensate for that dependency somewhere else, for example through longer contracts or higher switching costs for the customer.
The basis is a breakdown of revenue by origin: contractually fixed, habitual behaviour without a contract, and one-off. Then, per category, you count the share of work needed to retain that revenue, and who or what carries out that work.
A workable sequence:
1. Map the renewal rate of the last two to three years per customer segment, not as a single nationwide average but per type of customer. 2. Set next to each revenue stream the tasks needed to retain that customer: follow-up, invoicing, providing service, renewing the contract. 3. Mark for each task whether it is currently done by a system, by a system with human approval, or entirely by a person. 4. Determine what happens to the recurring revenue if that person leaves. Does the process keep running, or does it stall.
This fourth step is exactly where owner and team dependency comes into play: recurring revenue that looks solid on paper can still receive a low score from a buyer once it becomes clear that one person is the glue. Conversely, a business with less contractual certainty can still score well if the retention process itself is robust, visible in processes and systems that do not run on a single workstation.
Improvement here does not lie in relabelling revenue, but in shrinking the manual portion of customer retention and recording evidence of that.
Three directions that make a difference in practice:
This is not about personnel decisions. What happens to a role or an employee when a task disappears or changes falls under the employer's own legal requirements and obligations. The point here is mapping which work keeps a process running, independent of who carries out that work today.
The question of whether recurring revenue in your business rests on a person or on a process can already be partly answered with a few targeted questions. The free value check consists of eight short questions, one per value driver, and gives a picture of which driver is putting the most pressure on your price today. The full value scan, with evidence per driver and a two-year calendar towards the exit moment, is under construction.